A 13-slide research deck produced jointly by Google and Boston Consulting Group, titled "BCG-Google The New Era of Marketing Partnerships" and subtitled "2026 research results and findings", reached the wider media buying community on August 21, 2026, when Deepjyot Bakshi, Account Director for Programmatic at JCDecaux UK, posted a link to the file on LinkedIn. The document itself is hosted on bcg.com. Its underlying survey was fielded in January 2026.

The findings describe a marketing services market in which the incumbent agency structure ranks at the bottom of every shortlist marketers name for the work they say they are buying next.

A survey of 990, a dataset of 387

The methodology note on the second slide states the research covered nearly 1,000 respondents, given precisely as 990, across 10 countries. The country list runs across three regions: the United States, Canada and Brazil in the Americas; the United Kingdom, Germany and France in EMEA; and Japan, India, Indonesia and Australia in Asia-Pacific.

Every data slide that follows, however, carries the same source line: Google/BCG, The New Era of Marketing Partnerships: Marketer Survey, January 2026, Global, N=387 Marketers. The remaining respondents, roughly 600 of them, are not broken out anywhere in the circulated version. The deck offers no cross-tabulation by country, by category, by budget size or by respondent seniority.

Four questions frame the exercise, according to the deck: what marketers need in a landscape disrupted by AI, which partner types are becoming more relevant, where investment is shifting, and how agencies and partners can grow as those dynamics change.

Eighty-one per cent want it. Forty-five per cent have it

The first substantive finding is a gap. According to the survey, 81% of marketers want partners to provide integrated capabilities across creative, media, earned and commerce. The figure derives from a seven-point importance scale, with scores of five through seven counted as agreement. Asked to describe their current level of integration across those same four areas, 45% say they are integrated.

That 36-point spread is the number the deck builds its argument on. It also sets up the sharpest slide in the set, headed "There are major disconnects between what marketers and agencies say and do."

On the marketer side, the slide contrasts the stated 81% preference with the observation that procurement teams still issue separate requests for proposal and create fragmentation. On the agency side, it contrasts the pitch line "We offer end-to-end orchestration across your full marketing ecosystem" with the note that ground teams are channel specialists who have never been in the same room.

The agency-side statements carry no survey citation. The source line on that slide covers only the 81% marketer figure and the question that produced it. The characterisation of agency behaviour is presented as observation, not measurement.

Reach is a tablestake. Innovation moves budget

A slide headed "The basics are basic - innovation drives movement in media budgets" ranks the factors marketers say influence channel allocation now and expect to influence it in two years.

Audience reach and scale leads at 80%. Historical measurement and marketing mix modelling data follows at 70%. The deck files both under a heading reading "Tablestakes to be in the game", a classification that says something about how the two most cited factors are read: as entry conditions rather than differentiators.

The movement sits elsewhere. Innovation and new platform capabilities is expected to reach 47% within two years, an increase of 23 percentage points and the largest shift recorded on the chart. Creative strategy and asset availability registers 54%, competitive activity and category norms 46%, leadership mandates and internal priorities 42%, agency recommendations 36% and pricing or incentives 24%.

Agency recommendations, in other words, rank sixth of eight among the factors marketers say shape where their channel money goes.

Non-working spend moves toward AI

The deck's fourth slide addresses non-working spend, the portion of a marketing budget that pays for services, tools and infrastructure rather than for media inventory. Respondents were asked how that spend is allocated now against the next two years across four buckets: AI solutions, insights and analytics, data and martech and workflow tools, and core creative and media services.

The definitions matter for anyone reading the categories against a media plan. AI solutions, per the footnotes, covers systems that autonomously plan and execute marketing workflows, including agentic solutions. Insights and analytics covers campaign measurement, reporting and experimentation. Data, martech and workflow tools covers data collection, enrichment, modelling, governance and infrastructure alongside platforms that automate marketing workflows. Core services covers media planning and buying, creative services and specialty services such as influencer marketing.

The direction is marked on the chart itself: a nine percentage point decline for core creative and media services and a nine point increase for AI solutions.

Alongside that reallocation, the deck reports the share of marketers anticipating increased investment in four specific areas. AI-driven creative development and production leads at 86%. AI-driven discovery follows at 84%. Digital commerce and agentic buying registers 71%. Influencer and creator marketing registers 55%.

A separate slide addresses where the money for those areas comes from. Paid media is identified as the legacy source. Owned and earned budgets, revenue growth management and commerce budgets, and product innovation budgets are identified as new funding sources feeding AI discovery, agentic commerce and creator marketing. The implication is that these categories are not purely a reallocation within existing media lines, and that budget authority for them sits partly outside the marketing function that agencies traditionally sell into.

AI platforms take three of four shortlists

The most direct commercial finding is a table of partner preference by opportunity area, drawn from a question asking which partner types respondents would consider for each. Multiple selections were permitted.

For AI-driven creative development and production, AI and agentic platforms lead at 58%, followed by independent agencies at 53%, martech and creative tool providers at 51%, performance agencies at 37% and holding company owned agencies at 36%.

For AI-driven discovery, AI and agentic platforms lead at 63%, followed by search specialists at 49%, consulting firms at 42%, holding company media agencies at 37% and holding company creative agencies at 25%.

For digital commerce and agentic buying, AI and agentic platforms lead at 50%, followed by commerce technology and retail media platforms at 42%, commerce specialists at 37%, performance agencies at 34%, consulting firms at 34% and holding company media agencies at 30%.

For influencer and creator marketing, the pattern breaks. Independent influencer agencies lead at 57%, followed by creator marketing tools and platforms at 44%, independent creative agencies at 39%, independent media agencies at 32% and holding company media agencies at 26%. AI and agentic platforms do not appear in the list at all.

Read across the four charts, a holding company agency occupies the bottom position in every one. The deck's own headline for the slide states that new types of partners are taking share and disrupting the landscape as marketers' needs change. The survey question, however, measures consideration rather than spend, and the data captures a single point in time. No prior wave is published for comparison.

What marketers say AI maturity means

Asked to name the most important people, process and technology considerations when assessing a partner on AI maturity, respondents produced a ranking led by AI expertise at 91%. Tool integration and interoperability follows at 77%, AI integration across processes at 75%, customisation at 72% and ability to scale at 71%.

The lower half of the list is where the readings get interesting. Day-to-day delivery use registers 67%, repeatable AI workflows 63%, differentiation and innovation 63%, accountability for AI decisions 56% and AI training and upskilling 53%.

The percentages sum well past 100 because the question asked for three selections in each of three categories, up to nine picks across ten options. Accountability for AI decisions ranking ninth of ten is nonetheless notable set against evidence from elsewhere in the market. Survey work published on August 20, 2026 found only 6% of marketers act on in-platform AI recommendations, with the restriction tracking accountability structures rather than model quality.

The headline number on the same slide is the one most likely to be quoted in agency new business decks: 97% of marketers indicate they would increase spend with a partner who delivers across the new opportunity spaces, defined as AI discovery, agentic commerce, AI creative and creator marketing.

That figure carries an obvious caveat. Stated willingness to increase spend, conditional on a partner delivering across four capability areas that few partners currently deliver across, is not a budget commitment. Research released on August 5, 2026 found only 53% of marketers get meaningful ROI from AI even among enterprise brands already deploying it at scale, while 89% were raising AI budgets regardless.

Two-thirds want to be billed on outcomes

Asked how they would prefer to compensate a marketing services partner for delivering the expansion opportunities they rated, respondents ranked outcome-based or performance-linked fees first at 66%. Hybrid models follow at 56% and project-based fees at 43%. Usage-based pricing registers 27%, revenue or profit share 26%, higher fixed fee or retainer 26% and SaaS or technology licensing fees 21%.

The deck does not present that as settled. A panel on the same slide, sourced to agency interviews and customer calls conducted across 2025 and 2026 rather than to the survey, lists three obstacles to outcome-based models: measurement and transparency, margin pressure, and client risk tolerance.

That matches what listed agency groups have been telling investors. WPP chief executive Cindy Rose told analysts that outcome-based pay remains years away, with a return to positive organic growth the nearer priority. The blocking constraint is the same one the deck names first: without an agreed definition of the outcome, there is nothing to price against.

Fixed retainers, the model that funds most holding company delivery, rank sixth of seven in marketer preference.

The mandate, and the closing pitch

The penultimate slide states the conclusion directly: "The marketer mandate is clear: holistic thinking." Three supporting lines describe building stronger internal partnerships to connect across silos, deepening external partnerships across agencies, media and technology providers, and designing what the deck calls human-to-agentic partnerships to accelerate the impact of AI. A circular diagram places paid, owned, earned and commerce around a centre labelled marketer outcomes, ringed in turn by agencies, tech and media.

The final content slide sets out four areas where the deck argues agencies and partners can win: building connectivity between specialised capabilities inside and outside the company, removing internal walls and friction from the operating model, exploring flexible delivery and commercial structures aligned to performance and outcome, and focusing on what it terms the expansionary thesis, looking beyond paid media for new value pools.

That closing argument is a recommendation to agencies from a research programme co-authored by a platform that sells media, measurement and creative tooling directly to the same advertisers. The deck does not address the conflict.

The DOOH reading the deck does not contain

The post that carried the file into the media buying feed on August 21 offered a specific interpretation. "Innovation is not about introducing new technology. It's about creating new value," Bakshi wrote. "That's why capabilities such as DCO, audience-led activation, retailer data integration, custom triggers and AI-powered planning are becoming increasingly important in Programmatic DOOH."

The reading is defensible against the innovation finding. It is not, however, drawn from the deck. Neither digital out-of-home nor dynamic creative optimisation appears anywhere in the 13 slides. The document names no channels at all beyond the four-way split of paid, owned, earned and commerce, and no vendors beyond the two organisations on the cover. The post carried nine reactions at capture.

What the numbers leave out

Several gaps limit what the research can support. The 990-respondent figure and the 387-marketer base are never reconciled. No agency-side survey data is published, despite the deck making claims about agency behaviour. No wave-on-wave comparison exists, so every directional statement about partners taking share rests on stated future intent rather than measured movement. Consideration is measured; spend is not.

The context around those gaps is documented. Research published in May 2026 found advertisers using AI for optimisation at 73% against 25% for creative production, a split that sits awkwardly beside the 86% who told Google and BCG they expect to increase investment in AI-driven creative development. Work published in July 2026 found generative AI lifted creative volume for 88% of marketers but quality for only 45%. Mediaocean's mid-year report identified AI media as the fastest-growing category at 60% while data gaps and stack integration remained the binding constraints on deployment.

Why it matters for media buyers

For anyone selling media services, the operative number is not 97% or 81%. It is 36%, 30%, 26% and 25%: the positions holding company agencies occupy at the bottom of each of the four opportunity charts.

That ranking arrives during a period of visible restructuring on the holding company side. LiveRamp shareholders approved a Publicis Groupe takeover on August 17, 2026, moving identity and clean room infrastructure inside an agency group. Platform consolidation is running in parallel, with Amazon DSP moving to a single global login across 34 markets and raising the question of what the agency line item justifies when the platform selects deals, connects data and proposes the plan.

For programmatic buyers specifically, the finding that agency recommendations rank sixth of eight in channel allocation decisions is a direct measurement of advisory influence. For anyone in creator marketing, the one category where AI and agentic platforms do not appear on the shortlist is worth noting against a United States creator economy projected at 43.9 billion dollars in 2026. And for procurement teams, the deck's own diagnosis is that separate requests for proposal are producing the fragmentation the same organisations say they want removed.

The research does not measure whether any of that changes hands. It measures what 387 marketers said in January 2026 about who they would consider.

Timeline

Summary

Who: Google and Boston Consulting Group produced the research. The published dataset covers 387 marketers drawn from a total of 990 respondents. Deepjyot Bakshi, Account Director for Programmatic at JCDecaux UK, circulated the file on LinkedIn.

What: A 13-slide deck titled "BCG-Google The New Era of Marketing Partnerships" reporting that 81% of marketers want integrated capabilities across creative, media, earned and commerce while 45% have them, that AI and agentic platforms lead partner consideration in three of four growth areas, that holding company agencies rank last in all four, that 97% would increase spend with a partner delivering across those areas, and that 66% prefer outcome-based or performance-linked compensation.

When: The survey was fielded in January 2026. The deck is dated 2026 and carries the subtitle "2026 research results and findings". It reached the media buying community through a LinkedIn post on August 21, 2026.

Where: Fieldwork covered 10 countries: the United States, Canada and Brazil; the United Kingdom, Germany and France; and Japan, India, Indonesia and Australia. The document is hosted on bcg.com.

Why: The research maps where marketing services budget is moving as non-working spend shifts from core creative and media services toward AI solutions, and it records which partner types marketers would consider for AI discovery, agentic commerce, AI creative and creator marketing. For agencies, the ranking places holding company structures at the bottom of every list. For advertisers and procurement teams, it quantifies a gap between the integration they say they want and the fragmented buying processes the same research says they operate.